1

Climate Change and Economic Growth in Africa: An Econometric Analysis

Ayodele F. Odusola
United Nations Development Programme
Babatunde O Abidoye
University of Pretoria




Abstract
The economic landscape of most Africa countries depends essentially on the dynamics of climate
change. Key sectors driving their economic performance and livelihoods such as agriculture,
forestry, energy, tourism, coastal and water resources are highly vulnerable to climate change.
This paper examines the empirical linkage between economic growth and climate change in
Africa. Using annual data for 34 countries from 1961 to 2009, we find a negative impact of
climate change on economic growth in Africa. Our results show that a 1 degree Celsius increase
in temperature reduces GDP growth by 0.27 percentage point for the region. A higher impact of
0.41 percentage point was however observed when the sample period was reduced to 1961 to
2000 indicating a reduction in the influence possibly given increase in efforts towards adapting
to climate change. The two largest economies in the Sub-Saharan Africa (South Africa and
Nigeria) played some significant role in ameliorating the negative economic impact of climate
change in the region.
Some policy options emerged from this study. First, mainstreaming climate change adaptation
into National Development Strategy and budgets could promote proactive engagement on the
formulation and implementation of climate change adaptation strategy. Second, the potential of
regional or multiple countries approach to climate change adaptation is high due to possibility of
economies of scale. Third, the role of South Africa and Nigeria in cushioning the negative
impact of climate change on other African countries tends to suggest the benefit of regional
integration in addressing this challenge.
2


Introduction
Climate change has been identified as one of the most daunting challenges facing the world in
this century and it is particularly more serious in developing countries largely due to their
geographic exposure, low incomes, greater reliance on climate-sensitive sectors and weak
capacity to adapt to the changing climate
1
. In fact, the economic landscape of most African
countries depends essentially on the dynamics of climate change. In Africa, the vulnerability of
the overall economy and key sectors driving economic performance such as agriculture, forestry,
energy, tourism, coastal and water resources to climate change has been acknowledged to be
substantial
2
. The geographical location of most African countries on the lower latitudes has
already put the region at a disadvantage where about 80 percent of damages from climate change
are concentrated. Any further warming would seriously affect productivity (Mendelsohn, 2009).
Yet, Africa contributes a small proportion to the global green hose emissions. As articulated by
Earth Trends (2009), it is less than 5 percent of total carbon dioxide-equivalent emissions and
this share is unlikely to grow substantially in the nearest future.
Over the past five decades (1960-2009), many countries in Africa (e.g. Sudan, Chad, Uganda,
Botswana and Tunisia) have experienced substantial rise in temperature ranging from 1
o
to
over 3
o
Celsius. The increasing knowledge that the continent contributes least to carbon finger
print but experiences the most severe impact of climate change provides incentives for Africa to
understand the costs of climate change to its economy and development prospects with a view to
informing policy decisions. This is not only as a result of losses to the economy that might be
linked to reduced agricultural productivity and labour losses but also from increases in
morbidity, mortality and social instabilities. These indirect impacts such as death and disabilities
associated with climate change have irreversible economic and welfare consequences. When
countries spend some resources to adapt to climate change, they incur opportunity costs of not
spending it on research and development and capital investment (e.g. infrastructure) that is a
binding constraint to growth and development.
However, there is limited empirical analysis on the damaging effects of climate change on the
African economy both collectively and at individual country levels. Due to dearth of literature on
this issue in the continent, there is yet to be a convergence on the magnitude of its impact on
economic growth both at the regional and country specific levels. This paper aims at quantifying
the implications of climate change on economic growth in Africa. Specifically the paper seeks to
answer the following questions: Does temperature matter in predicting economic growth in

1
Climate change manifests itself with temperature increases, changes in precipitation, rise in sea levels
thereby increasing the intensity of such natural hazards as storms, floods and droughts. For detailed analysis
of the various dimensions of climate change, their severity and implications on Africas development see !CC
"2##$%.
2
&ee 'ell et al "2#11% for the economy(wide impact and Boko, et al (2007) for sector specific effects.
)

Africa? Is there heterogeneity in the impact of climate change on the economic growth of
African countries? And what other factors are important in determining differences in growth
rates across the selected countries?
The outline of the paper is as follows. Section 2 touches on linkages between climate change and
economic growth in general and frames our paper in the context of other papers in the literature.
Section 3 presents the model and how our parameters of interest are estimated while Section 4
describes the data and analysis of the results obtained. The paper concludes with a summary in
Section 5.


2. What does the literature say about the link between climate change and economic
growth?
The literature is replete with the potential ways through which temperature could affect
economic activity. The damaging effect of changes in temperature on growth rate of GDP is
informed by both theoretic and empirical evidences. First, the destruction of ecosystems from
erosion, flood and drought, the extinction of endangered species and deaths resulting from
extreme weathers cause permanent damages to economic growth. Second, the resources required
to counter the impact of warming would reduce investment in economic and physical
infrastructures, research and development and human capital thereby reducing growth (Pindyck,
2011; Ali, 2012).
Theoretically, the linkage could be established through macroeconomic and microeconomic
dimensions. From the macroeconomic side, influence on the level of output such as agricultural
yields and economys ability to grow (for example by affecting investments or institutions that
influence productivity growth) are the two areas that are most emphasized (Dell, Jones, and
Olken 2012). From the microeconomic analysis dimension, the linkage include an array of
factors such as physical and cognitive labor productivity, conflict, and health, all of which could
have economy-wide implications (IPCC, 2007 and Gallup, 1999). For instance, increased
temperature leads to political instability, which in turn may impede factor accumulation and
productivity growth. To this end, this review provides an overview of the direct effect on
economic growth and the indirect effect on other variables such as morbidity and mortality.

Evidence from a panel of 136 countries over the period 1950-2003, (Dell, Jones, and Olken
2012) find three primary results from their study. First, higher temperatures substantially reduce
economic growth in poor countries. For instance, a 1
o
celsius rise in temperature in a given year
reduces economic growth by 1.3 percentage points on average. Second, higher temperatures
appear to reduce growth rates, not just the level of output. Third, higher temperatures have wide-
ranging effects, reducing agricultural output, industrial output, and political stability.

Bernauer, et al (2010), using global data for 1950-2004, observe that the impact of climate
change on economic growth is not robust to changes in climate change indicators and samples.
*

However, the moving average-based measure of temperature for Africa is associated with
negative effects though only at 10 percent level. Also Ali (2012), using a co-integration
analysis on Ethiopia finds a negative effect on growth. He specifically observed that change in
rainfall magnitude and variability has a long term drag-effect on growth.

Frankhauser and Tol (2005) provide theoretical and empirical investigations on the link between
climate change and economic growth using a simple climate-economy simulation model. They
argue that the capital accumulation effect is important, especially if technological change is
endogenous, and may be larger than the direct impact of climate change. The savings effect is
less pronounced. The dynamic effects are more important, relative to the direct effects. They
conclude that in the long run, for high direct impacts, climate change may indeed reverse
economic growth and per capita income may fall. For global warming of 3
o
C, the direct
damages to the economy are estimated to at least 15 percent of GDP. When the effect of capital
accumulation and peoples propensity to save are factored into the damages, the impact would be
higher.

Higher growing temperature can significantly affect agricultural productivity, farm income and
food security. The effect differs across temperate and tropical areas. In mid and high latitudes,
the suitability and productivity of crops are projected to increase and extend northwards while
the opposite holds for most countries in tropical regions (Gornall et al 2010). They find that a 2
o

Celsius rise in temperature in mid and high latitudes could increase wheat production by about
10 percent while in low latitude regions, it could reduce by the same amount. Their projection,
taking the effect of technology into account, found that rising temperature in Russia Federation
could increase wheat yield by between 37 and 101 percent by 2050s.

In addition, Salvador, et al (2004) find the effect of rising temperature on agriculture to be more
severe in Sub-Saharan Africa than other developing countries. Results from simulation exercises
suggest that if the climatic conditions (rainfall and temperatures) had remained at their pre 1960s
level, the gap of agricultural production between Sub-Saharan Africa and other developing
countries at the end of the 20th century would have been only 32 per cent of the current deficit.
Evidence from Ayinde et al (2011), using econometric analysis on Nigeria (1980-2005), reveals
that temperature change generated negative effect while rainfall change exerted positive effect on
agricultural productivity.

The Fourth Assessment Report of the IPCC provides some illumination results about the impact
of climate change on African development. For instance, projected reductions in yields in some
countries could be as much as 50% by 2020, and crop net revenues could fall by as much as 90%
by 2100, with small-farm holders being the most affected. It will also aggravate the water stress
currently faced by some countries - about 25% of Africas population (about 200 million people)
currently experience high water stress. The population at risk of increased water stress in Africa
is projected to be between 350-600 million by 2050 while between 25 and 40 percent of mammal
species in national parks in sub-Saharan Africa will become endangered (Boko, et al, 2007).

The survival of mosquito and malaria parasites are highly sensitive to daily and seasonal
temperature patterns. Evidence from Science Daily (2010) reveals that over the past four
decades, the spread of malaria to highland areas of East Africa, Indonesia, Afghanistan, and
+

elsewhere has been linked to climate change. This was a rare phenomenon in the cooler highland
areas about 50 years ago. Tanser et al (2003) also projected that due to changing temperature
pattern in Africa; there would be 57 per cent potential increase (mainly altitudinal) in malaria
distribution with surprisingly little increase in the latitudinal extents of the disease by 2100.
Boko et al (2007) also provide some insights into the climate change implications on public
health in Africa. As argued by Gallup et al. (1999), vector-borne diseases, particularly malaria,
can have such a large effect on labour productivity which could make many countries in Sub-
Saharan Africa to be trapped in a vicious cycle of diseaselow productivitypovertydeficient
health care. This has implications on the future welfare of the society.

Evidence from Rabassa et al (2012) reveals that weather shocks exacerbate child morbidity and
mortality in Nigeria rural areas and is of considerable magnitude. Rainfall shocks have a
statistically significant and robust impact on child health in the short run for both weight-for-
height and height-for-age, and the incidence of diarrhea. The fact that diarrhea is the leading
cause of child malnutrition, and the second leading cause of death for young children in the
country underscores the severity of the impact on human development and long term welfare
implications of weather changes. The intensity is highest in hottest regions. However, children
seem to catch up with their cohort rapidly after experiencing a shock.
In summary, climate change has negative impact in most tropical regions economies both
directly and indirectly. This is particularly important because of heavy reliance on rain-fed
agriculture which is the main livelihood of the largest segment of the population. To this end,
rising trend of temperature could have significant effect on agricultural productivity, farm
income and food security as well as indirect effect on labour productivity through impact on
public health.

3. Analytical framework for establishing the linkage
This section examines the standard cross-country growth models that can be used to estimate the
relationship economic growth and its key determinants. This is then used to specify the impact of
omitted variable bias on parameters of interest.
3.1 The Basic Cross-Country Growth Regression Model
Following the framework in Barro (1991), Levine and Renelt (1992) and Sala-i-Martin (1997b),
we model y

, economic growth of country i, as follows:

y

= y
0
+ z

y
k
+ [x

+ e

iiJ
~
N(u, o
s
2
)

,

In the above, y

denotes the average growth rate of GDP of country i over a certain year range. In
line with Levine and Renelt (1992), z

denotes a vector of explanatory variables of country i over

the same year range that are believed to influence growth and will include a set of variables that
are always included in the regression, and then a subset of variables chosen from a pool of
variables identified by past studies as potentially important explanatory variables of growth. x

is
a certain variable of interest potentially important explanatory variables of growth.
3

The cross-country growth regression model differs in an important way from models that use
panel data such as Savvides (1995) and Hoeffler (2002). These models that incorporate panel
data tend to address some issues that single cross-country regressions may have. Some of these
issues as pointed out in Hoeffler (2002) include the issue of reducing the time series to a single
(average) observation; omitted variable bias issue and endogeneity of some of the regressors.
Also, these models are used to capture country-specific effects. However, some of these issues
may not be as pronounced in the single cross-country regressions. For example, the bias of using
a single (average) observation may be small if the variable has not changed much over time as is
the case for some of the variables that are included in the economic growth literature.
4
Also,
endogeneity problem is usually addressed by using the initial values of the variables that may be
endogeneous in the model.
Attempt to solve the omitted variable bias has however led to an influx of variables that has been
included over time with the norm of looking at variables that are significant to determine the
factors that explain differences in growth rates across countries. This has led to the literature
addressing uncertainty in the variables to be included in these models. Levine and Renelt (1992),
Sala-i-Martin (1997a and b), and Fernandez, Ley, and Steel (2001) all investigated the issue of
model uncertainty. Fernandez, Ley, and Steel (2001) used a Bayesian framework that allowed
them to deal with both model and parameter uncertainty using Bayesian Model Averaging.
Ignoring the issue of using averages, the single cross-section growth regression specification
appropriately models differences in growth patterns of countries when there is no correlation
between the variable of interest and other explanatory variable. However, when the variable of
interest is potentially correlated with unobserved variables, the single cross-section growth
regression specification will lead to inconsistent estimate of the variable of interest. In the
following section, we describe a Bayesian estimation algorithm which properly accounts for the
impact of correlation between unobserved variables and the outcome of interest. This
specification is important for us to study the impact of climate change on economic growth.



)
-ypically, the estimation involves varying the pool of potentially important e.planatory variables of growth.
*
t can be argued that variables such as school enrolment, population growth and labor force has not
significantly diverged from the norm over a span of the sample period used in many of the growth studies.
$

3.2 Linear Hierarchical Model
Using Bayesian approach, this paper first assumes that climate change variables such as
temperature will have a different impact on GDP across countries and should be permitted to
vary across countries. There is however a degree of commonality across the continent on its
impact drought in South Africa will have an impact on the economy of neighboring countries
even if it was not as severe as that of South Africa. On the other hand, climate variables may also
have an impact on many of the explanatory variables that may be included (observed) or not
included (unobserved) in the regression equation. Consistent estimate of the parameters of
temperature and observed explanatory variables such as initial GDP per capita will require that
these variables be uncorrelated with the unobserved variables. This condition is unlikely to hold
especially given unavailability of data for many of the variables that can potentially influence
economic growth and related to temperature. This is the classic omitted variables bias and
inconsistency problem.
5

We propose a linear hierarchical model that is similar to the non-Bayesian fixed effects model
but exploits the hierarchical prior framework to estimate the parameters of the observed variables
that influence economic growth. The model proposed is in the spirit of the normal hierarchical
linear model described in Lindley and Smith (1972) and is similar to the model in
Abidoye, Herriges, and Tobias (2012) controlling for observed and unobserved variables using
country specific constants.
6
In particular, we will introduce a country-specific constant term that
captures both the observed explanatory variable and unobserved explanatory variable.
That is, we can employ the model:
y
t
= o

+ x
t
[

+ e
t
i = 1,2, , N; t = 1,2, , I
Where
o

= y
0
+ z

o
y
k
+ z

u

This problem resolves the omitted variable bias since e
t
is no longer correlated with the variable
of interest (x
t
). However, the impact of the observed explanatory variables on economic growth
will not be separately identified in the classic fixed effects specification.
We will estimate the above equation in a Bayesian framework and will adopt the blocking
strategy in Abidoye, Herriges, and Tobias (2012) by proceeding in a manner that is similar to the

+
Abidoye, /erriges, and -obias "2#12% illustrate this problem in a 0andom 1tility 2a.imization setting but
the setting is similar to ours by replacing choice alternatives with time.
,
'etailed description of this model and similar hierarchical models in the 3ayesian framewor4 can be found
in 5oop, !oirier, and -obias "2##$%.
6

classic fixed effects model by isolating the impact of the unobservable (capturing them entirely
in the country- specific constants) and insulate the climate parameter from their effects.
7


3.3 Hierarchical Priors
As stated in the previous section, the country-specific constants capture explanatory variables
that are included and not included in the regression that might explain the differences in
economic growth rates across countries.
8
The interactions of all country level variables that are
not of interest but typically included in cross-country growth models are solely captured in the
country-specific constants. We are also interested in estimating the correlation between the
climate variable and the unobserved variables that may not be captured in the regression. This
correlation will indicate the impact increase in temperature will have on these variables.
In our Bayesian approach, we capture the above by introducing a hierarchical structure into our
model, by assuming that each country shares some degree of commonality in their temperature
and economic growth by assuming that the country-specific constant and parameter on
temperature are drawn from the same normal population. That is, we allow for correlation
between the impact of temperature and other factors that may influence economic growth.
Specifically:

0

= j
o

[ ~ N(0
0
,) . (2)
Where:
0
0
= _
y
0
+z

o
y
k
[
0
_ = j
z

y
[
0
[
=_
_
uu
_
u[
_
u[
_
[[
_ = _
o
u
2
po
u
o
[
po
u
o
[
o
[
2
_
z

includes a constant term and the observed/included explanatory variables that influence
growth in country i. The correlation between temperature and the intercept is captured with p.
There are some silent features of our model that is worth mentioning our specification, as is the
case with most cross-country growth model will not solve the problem of potential correlation
between the included explanatory variables and the unobserved variables. It is typically assumed

$
As is pointed out in Abidoye, Herriges, and Tobias (2012), this simply echoes standard result that the fixed effects
estimator is unbiased even when correlation exists between the fixed effects and other explanatory variables
included in the model.
8
Also, the interactions of all country level variables that are of interest but typically included in cross-country
growth models are solely captured in the country-specific constants.
7

that this assumption holds. However, if this assumption does not hold, our specification can be
extended to make use of instrumental variables approach to consistently estimate y. In this paper,
we are particularly interested in consistently estimating [

and [
0
. Even when such correlation
exists, the inclusion of country-specific constants and our posterior simulator will yield
consistent estimates of the parameters of interest.
To complete our model, we specify priors for the remaining parameters. These are enumerated
below:
y~ N(p
y
, I
y
)
[
0
~ N(p
[
, I
[
)

-1
~ w(|p
0
R]
-1
, p
0
)
o
s
2
~ I0(o
s
, b
s
)

The hyper-parameters of the priors above, such asp
y
, I
y
, p
0
, o
s
, b
s
e.t.c., are supplied by the
researcher and are in general chosen to be relatively vague to allow dominance of the
information from the data. The notation N refers to the normal distribution, whereas W(.,.)
represents a Wishart distribution and IG(.,.) represents the inverse gamma distribution. There are
parameterized as in Koop, Poirier, and Tobias (pp. 336-339). These particular families of priors
are chosen primarily because when combined with the likelihood function yield conditional
posterior distributions that are easily recognized and sampled. These proper priors also make
model comparison and calculation of Bayes Factor relatively easy. Our prior means p
y
anu p
[

are set to zero vectors with the respective variance I
y
anu I
[
set to identity matrix and 25
respectively. The priors (hyperparameters) on the variance term are also selected by choosing
o
s
= S and b
s
= 1(4u).
9
p
0
is set to be equal to 5 and the prior is chosen to reflect some degree
of variability in the temperature and economic growth across countries. All these priors are
chosen to be reasonably diffuse and non-informative.

3.4 The Posterior Simulator
We fit the model using the Gibbs sampler and employ a number of blocking steps to mitigate
autocorrelations and consistently estimate our parameters of interest. Before describing these,
first let 1 = |{0

]
=1
n
y [
0

-1
o
s
2
] and define 1
-o
as all the elements of 1 other than .
The joint posterior distribution for all the parameters of this model can be written as:

7
-his chooses the prior mean for sigma82 e9ual to 2# with standard deviation also e9ual to 2#
1#

p(1|y)
__p(y

|H

, 0

, o
s
2
)p(0

|y, [
0
,
-1
)
N
=1
_ p(y|Z, p
y
, I
y
) p([
0
|X, p
[
, I
[
) p(o
s
2
|o
s
, b
s
) p(
-1
|p
0
, R)

Step 1: Draw {0

]
=1
n
|1
-{0
i
]
, y

This complete conditional is proportional to the joint posterior distribution p(1|y). Absorbing all
the terms that do not involve 0

into the normalizing constant of this condition gives us the

complete posterior conditional for 0

. We have stacked the observations over time for each

country so that:
y

= _
y
1
y
2
.
y
t
_, H

= _
1 x
1
1 x
2
. .
1 x
1
_.
Thus we obtain:
p(0

|1
-0
i
, y)~N(
0
i
J
0
i
,
0
i
), i = 1,2, , N
Where

0
i
= _
H

i
H

o
s
2
+
-1
_
-1
J
0
i
=

H

i
y

o
s
2
+
-1
0
0

We sample each of the 0

by sampling from the corresponding complete conditional.

Step 2: Complete Posterior Conditional for y
The complete posterior conditional for is can also be gotten as proportional to the joint
posterior distribution.
p(y|1
-y
; y) __p(0

|y, [
0
,
-1
)
N
=1
_ p(y|Z, p
y
, I
y
)
Once we condition on the 0

s, the mean of the y is simply the linear regression of the country-

specific constants on the variables of interest. That is, we can write:
_
o
1
o
2
.
o
N
_ = _
z
1
z
2
.
z
N
_ y + _
u
1
u
2
.
u
N
_
11

o = zy + u
Where the vai(u) = _
uu
- _
u[
_
[[
-1
_
u[
since it is a conditional distribution from 0

Thus we can write:
y|1
-y
; y~N(
y
J
y
,
y
)
Where

y
= (
z
i
z
vai(u)
+ I
y
)
-1

And
J
y
=
z
i
o
vai(u)
+ I
y
p
y

Step 3: Complete Posterior Conditional for [
0

The complete Posterior Conditional for [
0
is similar to that of y above. Once we condition on
the 0

s ([

's); we can write:

_
[
1
[
2
.
[
N
_ = _
1
1
.
1
_ [
0
+ _
:
1
:
2
.
:
N
_
[ = 1
N
[
0
+ :
Where the vai(:) = _
[[
- _
u[
_
uu
-1
_
u[


In this form, the posterior for [
0
will be:
[
0
|1
-[
0
; y~N(
[
0
J
[
0
,
[
0
)
Where

[
0
= (
N
vai(:)
+ I
[
0
)
-1

And
J
[
0
=
_ [

Nvai(:)
+ I
[
0
p
[
0

12


Step 4: Complete Posterior Conditional for o
s
2

o
s
2
|1
-c
s
2; y ~I0 _N -
I
2
+ o
s
, ju.S (y

- H

)
i
(y

- H

) + b
s
[
-1
]
Step 5: Complete Posterior Conditional for
-1

-1
|1
-
-1; y ~ w _j(0

- 0
0
) (0

- 0
0
)' + Rp
0
[
-1
, N + p
0
]

4. Data, estimation techniques, descriptive statistics and analysis of results
4.1 The Data
This section describes the data used to run the models specified above and the descriptive
analysis presented in the next section. Temperature data for each African country was gotten
through the Climate Research Unit (CRU). The study used observed gridded monthly mean
temperature data from the Climate Research Unit (CRU, version 3.0, Mitchell and Jones 2005).
The CRU dataset is based on station data and has a 0.5X0.5 resolution. The Global Gridded
Climatology data is presented at a new high resolution and made available by the Climate
Impacts LINK project, Climate Research Unit, University of East Anglia, Norwich, UK
(Mitchell and Jones, 2005). The Climatic Research Unit (CRU) data set is composed of monthly
0.50 latitude/longitude gridded series of climatic parameters over the period 1901-2009 however
the data used for this paper runs from 1961-2009.
For the purpose of studying the impact of climate change on economic growth in Africa, we find
it suitable to use data from the Africa Development Indicators (ADI) (2011) publication of The
World Bank. Economic growth is measured as the Annual percentage growth rate of GDP at
market prices based on constant local currency. Population data was also obtained from ADI.
Total population is based on the de facto definition of population, which counts all residents
regardless of legal status or citizenship--except for refugees not permanently settled in the
country of asylum, who are generally considered part of the population of their country of origin.
The values shown are midyear estimates.
Human capital investment is proxied for by primary school enrolment rates and life expectancy.
Although previous research (e.g. Mankiw, Romer, and Weil (1992) and Gemmell (1996)) has
argued that using the level of human capital with school enrolment can be problematic, we still
include it in the estimation. It has been used in many other studies and we let the model
likelihood dictate if it should be included or not.
1)

The Data is available for 34 countries. The sample consists of: Algeria, Benin, Botswana,
Burkina Faso, Burundi, Cameroon, Central African Republic, Chad, Congo, Dem. Rep., Congo,
Rep., Cote d'Ivoire, Egypt, Arab Rep., Gabon, Ghana, Kenya, Lesotho, Liberia, Madagascar,
Malawi, Morocco, Niger, Nigeria, Rwanda, Senegal, Sierra Leone, South Africa, Sudan,
Swaziland, Tanzania, Togo, Tunisia, Uganda, Zambia, and Zimbabwe.

4.3 Estimation and Testing
The algorithm described in Section 3 has been used to run our posterior simulator for 100 000
iterations discarding the first 5 000 of these as the burn-in. Results from these runs suggested that
the chain mixed reasonably well and appeared to converge within a few hundred iterations.
Although our point estimates are suggestive of good performance, any Makov Chan- Monte
Carlo (MCMC)-based inference can be affected by the degree of correlation among the
parameter draws over sequential iterations. The mixing of the posterior simulations has been
used to determine how many draws are needed to achieve the same level of numerical precision
that would be obtained under an independent identically distributed (iid) sampling. A high
degree of correlation will lead to a slow mixing that may not let us explore all areas of the
posterior as needed. These inefficiency factors, as they are called can be calculated by using the
definition of the numerical standard errors (NSE) of a Monte Carlo estimate with correlated
draws. The mean estimates can be obtained as:
NSE(0

m
) =
_
o
2
m
_1 + 2 _1 -
]
m
] p
]
,
m-1
]=1

Where 0 represents an arbitrary scalar parameter of interest, m denotes the number of post-
convergence simulations, 0

m
represents our estimate of E(0|y) as the sample average of our
post-convergence draws, p
]
represents the correlation between simulations ] periods (iterations)
apart and o
2
Io(0|y).
The NSEs for our models are extremely small relative to the mean estimates which strongly
indicate our simulation based estimates accurately approximate the posterior means of this
selection parameters. This, again, suggests that our algorithm mixes quite well.
10





1#
:e do not present the values for the ;&<s but all of them are less than #.##+.
1*

4.3 Descriptive Analysis
This section presents the main feature of temperature dynamics in the 34 African countries used
in this paper. -able 1 presents the minimum and maximum temperatures, the difference between
the minimum and maximum, the mean (1961 and 2009) and the absolute change between 1961
and 2009. Based on the mean value, Burkina Faso, Senegal, Benin, Niger and Ghana are among
the hottest countries in Africa while Lesotho, Morocco, South Africa, Rwanda and Tunisia
appear to be the coldest. Sudan, Botswana and Niger experienced the highest swings between
the minimum and the maximum temperature over the period of 49 years. Countries that changed
by more than 2
o
Celsius between 1961 and 2009 are Sudan (3.04), Chad (2.61), Niger (2.47) and
Egypt (2.15).
Figure 1 shows the trend of temperature for countries with the highest swings over the period.
Sudan and Chad have the highest levels and have been rising consistently between 1961 and
2009. They are followed by Uganda, Botswana and Tunisia. Countries that experienced some
relative stability in temperature between 1961 and 2009 include Madagascar, Congo Democratic
Republic, Gabon, Liberia and Sierra Leone (see Figure 2).
As shown in Figure ), lag of temperature change appears to have inverse relationship with the
change in current output. This is a clear indication that lag of change in temperature is a good
predictor of change in the level of outputs. A similar trend is observed for agriculture (Figure *).
The pattern for most countries follows the regional trend as shown for Sudan in Figure +. The
correlation index between temperature and agriculture value added is -0.61.
Table 1: Descriptive analysis of temperature (1961-2009)
Row Labels Min Max Max -
Min
Mean
Temperature
Standard
errors
Absolute
change in
temperature
(1961 2009)
Algeria 21.72 24.04 2.32 22.96 0.55 1.01
Benin 26.62 28.61 1.99 27.56 0.46 1.02
Botswana 20.39 23.21 2.82 21.86 0.62 1.46
Burkina Faso 27.54 29.12 1.58 28.32 0.39 1.34
Burundi 19.83 21.73 1.91 20.48 0.46 0.96
Cameroon 24.00 25.51 1.51 24.71 0.33 1.01
Central African
Republic
24.28 26.02 1.74 25.10 0.45 1.06
Chad 25.72 28.33 2.61 26.99 0.58 2.61
Congo, Dem. Rep. 23.79 25.33 1.54 24.62 0.30 0.64
Congo, Rep. 23.75 25.10 1.35 24.23 0.33 1.01
Cote d'Ivoire 25.58 27.17 1.59 26.41 0.32 0.21
Egypt, Arab Rep. 21.54 23.74 2.19 22.57 0.56 2.15
1+

Gabon 24.17 25.91 1.75 25.09 0.31 0.46
Ghana 26.45 28.14 1.70 27.29 0.37 0.68
Kenya 23.49 25.55 2.06 24.59 0.43 1.06
Lesotho 11.48 13.40 1.92 12.39 0.49 0.49
Liberia 24.71 26.10 1.39 25.38 0.29 0.42
Madagascar 21.67 22.81 1.14 22.30 0.32 0.05
Malawi 21.20 22.91 1.71 22.01 0.40 0.71
Morocco 16.04 18.47 2.43 17.36 0.53 0.29
Niger 26.20 28.68 2.47 27.45 0.49 2.47
Nigeria 26.19 27.84 1.65 26.93 0.38 1.52
Rwanda 18.32 20.24 1.92 18.99 0.48 1.09
Senegal 27.14 29.06 1.92 28.08 0.46 0.47
Sierra Leone 25.60 26.97 1.37 26.25 0.32 0.60
South Africa 16.96 18.60 1.64 17.85 0.42 0.82
Sudan 25.82 28.86 3.04 27.26 0.73 3.04
Swaziland 19.47 21.16 1.68 20.21 0.44 0.34
Tanzania 21.83 23.38 1.55 22.52 0.42 0.66
Togo 26.24 28.27 2.04 27.19 0.44 0.84
Tunisia 18.40 20.87 2.47 19.71 0.68 1.14
Uganda 22.01 24.58 2.57 23.00 0.67 1.90
Zambia 20.96 23.29 2.33 21.84 0.52 0.92
Zimbabwe 20.29 22.91 2.62 21.28 0.56 1.14

4.4 Analysis model of results
The analysis of the link between temperature and economic growth is based on the common
intercept
0
, common slope
0
, variance parameters of the second-stage covariance (denoted
by
2

and
2

), the correlation between the intercept and slope, denoted

,
, for all the selected
African countries based on data availability. In addition to the pooled result, we analyze the
slope and intercept results for 34 African countries. We report parameter posterior means and
posterior probabilities of the effect of temperature change being negative on economic growth
[denoted P (.<0|y)]. Multivariate regressions are also examined for the complete effect to be
manifested.
-able 2 presents the result of common parameter estimates. The results of the multivariate
regression are generally consistent with previous studies and will not be discussed at length.
Although evidence is not strong for population growth, investment and human capital (proxied
by net primary school enrolment and life expectancy) all contribute positively to economic
growth. The results show the importance of initial condition (the log of initial GDP per capita)
in the continent growth process. However, it does not provide evidence in favor of unconditional
1,

convergence.
11
The result shows that the correlation between the temperature and other factors
that influence economic growth is mostly negative and rather precisely estimated. This suggests
that countries with lower temperature increases will tend to have higher growth rates.
-able ) presents the results for the pooled and individual countries. For all countries, the
relationship between temperature and economic growth is largely negative. Evidence from the
shorter sample (1961-2000) tends to show higher level of damages to economic growth than the
larger sample. A 1
o
Celsius rise in temperature slows down economic growth by 0.41 percent for
the smaller sample with a probability value of 0.98. This implies that the chances that the effect
of temperature change on economic growth in Africa is negative are 98 percent most of the
times. For the larger sample, a 1
o
Celsius increase in temperature reduces GDP growth by 0.27
percent. This is better illustrated in Figure that shows the distribution of the pooled mean
effect of temperature on GDP growth in Africa. Majority of the posterior distribution for the
shorter sample is clearly massed away from zeros. However, although for the full sample, the
majority of the posterior distribution is still massed away from zeros, the evidence is not as
strong as in the shorter sample period. As could be gleaned from figure 6, the extended sample
size has reduced the mean effect of temperature changes on economic growth from about -0.5 for
the 1961- 2000 sample to -0.3 for the 1961-2009 sample.
The reduced influence on the larger sample could be as a result of adaptation programmes such
as use of drought resistant seedlings that is being adopted in many African countries. Recent
surveys especially from La Rovere et al (2010) reveal that dissemination and distribution of
drought resistance maize among many African countries have yielded some positive results in
terms of raising yields by 10 to 34 percent compared with non-drought resistant varieties
12
.
To gauge the impact of the four largest economies (South Africa, Egypt, Nigeria and Algeria) on
the overall impact on the pooled data, a with-or-without analysis reveals the strength of these
countries on the overall performance. When South Africa and Nigeria were removed, the
severity of the impact is higher both in parameter estimates and probability. A 1
o
Celsius rise in
temperature increases the damage effect from 0.27 (for all countries) to 0.35 percent with a
probability of 97 percent. Several factors could account for this significant influence on the

11
-his is not really new given that studies such as 3arro and &ala(i(2artin "1772% and
2an4iw, 0omer, and :eil "1772% have also reported failure of unconditional convergence when tested for
heterogeneous group of countries.
12
<vidence from the nternational nstitute of -ropical Agriculture "-A% reveals that the 'rought -olerance
2aize for Africa "'-2A%, a =oint initiative of the nternational 2aize and :heat mprovement Center
"C22>-% and the -A has led to dissemination of )* new drought(tolerant maize varieties to about 2
million small farm holders in 1) pro=ect countries?Angola, 3enin, <thiopia, @hana, 5enya, 2alawi, 2ali,
2ozambi9ue, ;igeria, -anzania, 1ganda, Aambia, and Aimbabwe?between 2##$ and 2#11.
httpBCCwww.iita.orgChome(news(
ssetDpEpEidF1#1E;&-A;C<E1n3&GpEpElifecycleF#GpEpEstateFnormalGpEpEmodeFviewGpEpEcolEidFcolumn(
2GpEpEcolEposF1GpEpEcolEcountF+GE1#1E;&-A;C<E1n3&EstrutsEactionFH2FassetEpublisherH2FviewEcont
entGE1#1E;&-A;C<E1n3&Eurl-itleFdrought(tolerant(maize(wins(2#12(u4(climate(change(
awardGE1#1E;&-A;C<E1n3&EtypeFcontentGredirectFH2FhomeI.1<Cumd$z'2
1$

continent. The most obvious is the level of economic integration of these two countries
especially Nigeria in ECOWAS and South Africa in SADC and COMESA. All the neighboring
countries to these large economies always benefit from their relaxed trade relations. When
Egypt and Algeria were removed from the larger sample, the results the severity declined as well
as the probability of occurrence (Table 3). The two countries are net importers of grains
especially wheat (the main staple food). Egypt, for instance, depends on her neighboring
countries such like Sudan and Ethiopia and has also acquired land for agricultural activities in
these countries especially Sudan.
Analysis of the individual country provides more illuminating results. It shows that countries in
Africa share some degree of commonality on the effect of temperature changes on GDP
growth rate. The intercept and slope parameters are drawn from the same normal population with
temperature having a negative impact on GDP growth rate in Africa. Across the 34 countries,
the effect of temperature on economic growth is largely negative with
i
ranging between -0.338
for Rwanda and -0.545 for Zambia (for the small sample) and 0.128 for Sudan and 0.495 for
Zimbabwe (for the large sample). As indicated in table 3 and using the large sample size (1961-
2009), climatic change will have the highest impact on countries such as Zimbabwe, Algeria,
Gabon, South Africa, Swaziland, Zambia, Tunisia, and Botswana. The least effect (although still
very high) is noted among countries such as Rwanda, Sudan, Chad and Uganda.
There is also the proximity effect on a few countries in terms of the similarity of the effects of
climate change on economic growth. Chad and Sudan; and South Africa, Lesotho and Swaziland
are good examples. An important policy implication of this is that there could be economies of
scale in dealing with the effect of climate change both in term of mitigation and adaptation
through cross border or regional efforts.

Conclusions
The vulnerability of the African economy and key sectors driving economic performance (such
as agriculture, forestry, energy, tourism, coastal and water resources) to climate change has been
acknowledged to be substantial. The inability of most African countries to create jobs in the
formal sectors of the economy could further strengthen the dependence of majority of the
population on these sensitive sectors. Yet, in the past five decades, many countries in Africa
such as Sudan, Chad, Uganda, Botswana and Tunisia have experienced substantial rise in
temperature ranging from 1
o
to over 3
o
Celsius. Managing the impact of climate change on
Africas economy has therefore become an important development challenge. This paper
examines the empirical linkage between economic growth and climate change in Africa.
Sudan, Botswana and Niger experienced the highest swings temperature variability. Their
temperature changed by more than 2
o
Celsius between 1961 and 2009 while countries
Madagascar, Congo Democratic Republic, Gabon, Liberia and Sierra Leone experienced some
16

relative stability. This study finds that lag of temperature change has inverse relationship with
the change in current output and appears to be is a good predictor of change in the level of
outputs. Based on data from 1961 and 2009, a 1
o
Celsius increase in temperature reduces GDP
growth by 0.27 percentage point. The impact is not homogenous across countries. The highest
impact is on countries such as Zimbabwe, Algeria, Gabon, South Africa, Swaziland, Zambia,
Tunisia, and Botswana while the least impact tends to be on Rwanda, Sudan, Chad and Uganda.
Given the critical role of agriculture in Africas economic growth and development, heavy
investment in research and development on the most appropriate adaptation interventions such as
development of drought resistant crops and promoting the development of water resources
management infrastructure (e.g. dams) would be vital in moving forward. To ensure a proactive
engagement in addressing this challenge, climate change adaptation should be integrated into
national development agenda and also reflected into budget implementation. The proximity
effect exhibited by the findings raises the need for economies of scale in dealing with the effect
of climate change. Sub-regional or cross border climate change mitigation and adaptation
initiatives may be more effective in the continent.
Using the four largest economies as the controlling factor for the impact of temperature changes
on economic growth provides some illuminating results with policy relevance. There is evidence
that Nigeria and South Africa serve as important stabilizer to the impact of climate change in the
continent when compared with Egypt and Algeria. One possible link for this stabilizing role
could be economic integration especially Nigeria in ECOWAS and South Africa in SADC
and COMESA. During period of serious economic downturns in most neighboring countries to
South Africa and Nigeria, cross border trade with Nigeria and South Africa tends to douse such
pressure. Efforts to strengthen regional trade and integration may be an important strategy to
indirectly ameliorate effects of of climate change in the continent.

Although the focus is on climate change, the result also underpins the importance of investment
and human capital development (especially net primary school completion and life expectancy
rate) in the growth process. While quality investment is needed to fast track the growth process,
investment in human capital (including primary and secondary school education as well as child
and maternal health) is critical to sustain growth in the long run.





17


Table 2: Dependent Variable is D! "ro#t$ rate usin" data from 1961-2009 (!(% &0'y) in parent$eses)
Explanatory Variables M1

M2

M3

M4

M5

M6
Temperature (Pooled impat on
!"ria#
$
%&26'(

$
%&235)

$%&23)

$
%&25%)

$
%&2633

$
%&25(4
(%&(25
5#

(%&(%(
)#

(%&(1(
3#

(%&(14
'#

(%&(24
(#

(%&(1(
6#

*onstant (&'61

%&2144

%&2%'3

%&16%5

%&1135

%&%(4(

(%&%13
3#

(%&415
6#

(%&432
)#

(%&434
%#

(%&454
'#

(%&2)%
4#

+o, -nitial ./P per apita

1&2))6

1&1))

%&''1)

%&644)

%&4(('

(%&%16
5#

(%&%31
2#

(%&1%1
)#

(%&2%4
5#

(%&2)%
4#

Population .ro0t1

%&53%)

%&42%'

%&33)'

%&3%45

(%&3%1
(#

(%&342
6#

(%&365
'#

(%&3'%
5#

-n2estment ./P ratio

%&)41(

%&636(

%&5661

(%&226
2#

(%&266
1#

(%&2'3
(#

Primary 31ool Enrolment (ln#

%&6%%3

%&5%%4

(%&261
'#

(%&2(6
%#

+i"e expetany (ln#

%&4223

(%&325
3#

3i,ma s4uare
45&52%
1

45&516
5

45&52%
3

45&512
3

45&521

45&513
1

3i,ma beta %&14%4

%&14(1

%&1452

%&1455

%&1451

%&1433

orrelation (r1o#
$
%&6'%(

$%&6'1

$
%&6)'2

$6'%2

$
%&6)2)

$%&6)4
lo,5p(y#6 (Model Mar,inal
+i7eli1ood# 345)&1

4'5(&3

4(51

465'&6

4664&5

461)&6

2#

Table (: )ountry *evel result - Dependent Variable is D! ro#t$ +ate

!ll sample Period (1(61$
2%%(# 1(61 $ 2%%%
8o0 +abels

betais P(& 9%:y# ;etais P(& 9%:y#
Pooled Mean -0.2661 0.92 -0.4180 0.98
!l,eria

$%&3552 %&'3 $%&5126 %&(1
;enin

$%&2%5' %&)4 $%&3533 %&(%
;ots0ana

$%&314' %&)( $%&3365 %&'1
;ur7ina <aso

$%&1'63 %&)3 $%&32'( %&(%
;urundi

$%&3165 %&'% $%&46(6 %&(%
*ameroon

$%&3%(2 %&'1 $%&44'' %&(%
*entral !"rian 8epubli

$%&324( %&'2 $%&45(2 %&(1
*1ad

$%&136( %&6) $%&42)3 %&(1
*on,o= /em& 8ep&

$%&2'15 %&)( $%&4)25 %&(1
*on,o= 8ep&

$%&2624 %&)) $%&415' %&(%
*ote d>-2oire

$%&2'4) %&'% $%&3(54 %&(%
E,ypt= !rab 8ep&

$%&23%' %&)4 $%&3)56 %&(%
.abon

$%&3)11 %&'4 $%&4'5( %&(1
.1ana

$%&1((4 %&)3 $%&3)62 %&')
?enya

$%&2(32 %&)( $%&4245 %&(%
+esot1o

$%&34(1 %&'2 $%&462) %&(%
+iberia

$%&2141 %&)4 $%&3)4( %&(%
Mada,asar

$%&2'%% %&)) $%&45(6 %&(%
Mala0i

$%&252) %&)5 $%&4%%6 %&(%
Moroo

$%&31)1 %&)( $%&4''% %&(%
@i,er

$%&2'(5 %&'1 $%&45%1 %&(2
@i,eria

$%&1')3 %&)2 $%&3)44 %&(%
80anda

$%&%646 %&5' $%&33'4 %&'1
3ene,al

$%&26)5 %&'% $%&4262 %&(%
3ierra +eone

$%&1(%% %&)2 $%&41(6 %&(2
3out1 !"ria

$%&3341 %&'% $%&5%3) %&(%
3udan

$%&12'% %&6) $%&35'3 %&(%
30aAiland

$%&3345 %&'% $%&44(5 %&(%
TanAania

$%&1(43 %&)1 $%&3'45 %&(%
To,o

$%&2)'5 %&'% $%&3')% %&(%
Tunisia

$%&332% %&'1 $%&4)32 %&(%
B,anda

$%&1422 %&66 $%&35') %&'4
Cambia

$%&34)( %&'2 $%&544( %&(2
Cimbab0e

$%&45(1 %&'' $%&41'4 %&(%

21

Table ,: -stimation +esults +emovin" at *east .ne of t$e *ar"est -conomies in /frica (1961-2009)%
*ountries
DPooledD
Mean P(E:y9%#


8emo2in, !l,eria and E,ypt $%&1(6' %&'5)

8emo2in, @i,eria and 3out1 !"ria $%&3546 %&(6(5

8emo2in, 3out1 !"ria $%&2)%4 %&(1()

8emo2in, @i,eria $%&2(44 %&(41(

8emo2in, E,ypt $%&25)( %&(156

8emo2in, !l,eria $%&23'2 %&(%%2

22

Figure 1B -emperature -rends for five of the 2ost Jolatile "/igh variance% Countries in Africa


15
1)
1(
21
23
25
2)
2(
31
1(611(631(651(6)1(6(1()11()31()51())1()(1('11('31('51(')1('(1((11((31((51(()1(((2%%12%%32%%52%%)2%%(
3udan Tunisia B,anda ;ots0ana *1ad +inear (3udan#
2)

0i"ure 2: Temperature Trends for five of t$e *east Volatile (*o#est variance) )ountries in /frica

15
1)
1(
21
23
25
2)
2(
1(61 1(63 1(65 1(6) 1(6( 1()1 1()3 1()5 1()) 1()( 1('1 1('3 1('5 1(') 1('( 1((1 1((3 1((5 1(() 1((( 2%%1 2%%3 2%%5 2%%) 2%%(
3ierra +eone Mada,asar .abon *on,o= /em& 8ep& +iberia
2*

0i"ure (: )$an"e in /vera"e D! ro#t$ and la" of temperature c$an"e
(1910-2009)


0i"ure ,: )$an"e in /vera"e )$an"e in /"riculture Value /dded and la" of
temperature c$an"e (1910-2009)



$%&6
$%&4
$%&2
%
%&2
%&4
%&6
%&'
$4
$3
$2
$1
%
1
2
3
4
5
1
(
'
%
1
(
'
2
1
(
'
4
1
(
'
6
1
(
'
'
1
(
(
%
1
(
(
2
1
(
(
4
1
(
(
6
1
(
(
'
2
%
%
%
2
%
%
2
2
%
%
4
2
%
%
6
2
%
%
'
1an,e in ,dp ,ro0t1 1an,e in tem
$%&6
$%&4
$%&2
%
%&2
%&4
%&6
%&'
$2&5
$2
$1&5
$1
$%&5
%
%&5
1
1&5
1
(
'
%
1
(
'
2
1
(
'
4
1
(
'
6
1
(
'
'
1
(
(
%
1
(
(
2
1
(
(
4
1
(
(
6
1
(
(
'
2
%
%
%
2
%
%
2
2
%
%
4
2
%
%
6
2
%
%
'
1an,ein !,riV! 1an,e in tem
2+

0i"ure 2: )$an"e in Temperature and c$an"e in D! (3udan)

$2%
$15
$1%
$5
%
5
1%
15
$1
$%&5
%
%&5
1
1&5
1
(
6
2
1
(
6
3
1
(
6
4
1
(
6
5
1
(
6
6
1
(
6
)
1
(
6
'
1
(
6
(
1
(
)
%
1
(
)
1
1
(
)
2
1
(
)
3
1
(
)
4
1
(
)
5
1
(
)
6
1
(
)
)
1
(
)
'
1
(
)
(
1
(
'
%
1
(
'
1
1
(
'
2
1
(
'
3
1
(
'
4
1
(
'
5
1
(
'
6
1
(
'
)
1
(
'
'
1
(
'
(
1
(
(
%
1
(
(
1
1
(
(
2
1
(
(
3
1
(
(
4
1
(
(
5
1
(
(
6
1
(
(
)
1
(
(
'
1
(
(
(
2
%
%
%
2
%
%
1
2
%
%
2
2
%
%
3
2
%
%
4
2
%
%
5
2
%
%
6
2
%
%
)
2
%
%
'
2
%
%
(
3udan 1an,e in temp 3udan 1an,e./P
2,

0i"ure 6: Distribution of t$e 4!ooled4 5ean -ffect of Temperature on D! ro#t$ in /frica





2$

References:

Abidoye, B. O., J. A. Herriges, and J. L. Tobias. 2012. Controlling for Observed and
Unobserved Site Characteristics in RUM Models of Recreation Demand. American Journal of
Agricultural Economics 94 (5): 10701093.
Ali, S. 2012. Climate Change and Economic Growth in a Rain-fed Economy: How Much Does
Rainfall Variability Cost Ethiopia? Available at SSRN 2018233.
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2018233.
Ali, Seid Nuru, Climate Change and Economic Growth in a Rain-Fed Economy: How Much
Does Rainfall Variability Cost Ethiopia? (February 8, 2012). Available at SSRN:
http://ssrn.com/abstract=2018233 or http://dx.doi.org/10.2139/ssrn.2018233
Ayinde, O. E., O. O. Ajewole, I. Ogunlade, and M. O. Adewumi. 2010. Empirical Analysis of
Agricultural Production And Climate Change: A Case Study of Nigeria. Journal of Sustainable
Development in Africa 12 (6).
Ayinde, O.E, M. Muchie and G. B. Olatunji (2011): Effect of Climate Change on Agricultural
Productivity in Nigeria: A Co-integration Model Approach, Journal Human Ecology, 35(3):
189-194 (2011).

Barro, R. J. 1991. Economic Growth in a Cross Section of Countries. The Quarterly Journal of
Economics 106 (2): 407443.
Barro, R. J., and X. Sala-i-Martin. 1992. Convergence. Journal of Political Economy: 223
251.
Boko, M., I. Niang, A. Nyong, C. Vogel, A. Githeko, M. Medany, B. Osman-Elasha, R. Tabo
and P. Yanda, 2007: Africa. Climate Change 2007: Impacts, Adaptation and Vulnerability.
Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental
Panel on Climate Change, M.L. Parry, O.F. Canziani, J.P. Palutikof, P.J. van der Linden and
C.E. Hanson, Eds., Cambridge University Press, Cambridge UK, 433-467.

Dell, M., B. F. Jones, and B. A. Olken. 2012. Temperature Shocks and Economic Growth:
Evidence from the Last Half Century. American Economic Journal: Macroeconomics 4 (3): 66
95.
Dell, Melissa, Benjamin F. Jones and Benjamin A. Olken (2011): Temperature Shocks and
Economic Growth: Evidence from the Last Half Century,
http://www.kellogg.northwestern.edu/faculty/jones-ben/htm/climatechange.pdf

26

Fankhauser, S., and R. SJ Tol. 2005. On Climate Change and Economic Growth. Resource and
Energy Economics 27 (1): 117.
Fankhauser, Samuel and Richard S.J. Tol (2005): On climate change and economic growth,
Resource and Energy Economics 27 (2005) 117.

Fernandez, C., E. Ley, and M. F. J. Steel. 2001. Model Uncertainty in Cross-country Growth
Regressions. Journal of Applied Econometrics 16 (5): 563576.
Gallup, J. L., and J. D. Sachs. 2000. Agriculture, Climate, and Technology: Why Are the
Tropics Falling Behind? American Journal of Agricultural Economics 82 (3): 731737.
Gallup, J. L., J. D. Sachs, and A. D. Mellinger. 1999. Geography And Economic Development.
International Regional Science Review 22 (2): 179232.
Gemmell, N. 1996. Evaluating The Impacts Of Human Capital Stocks And Accumulation On
Economic Growth: Some New Evidence. Oxford Bulletin of Economics and Statistics 58 (1):
928.
Gornall Jemma , Richard Betts, Eleanor Burke, Robin Clark, Joanne Camp, Kate Willett and
Andrew Wiltshire (2010), Implications of climate change for agricultural productivity in the
early twenty-first century, Philosophical Transactions of the Royal Society of biological
Sciences, Volume 365, No. 1554 2973-2989 , 27 September 2010
Hoeffler, A. 2002. The Augmented Solow Model and the African Growth Debate. Oxford
Bulletin of Economics and Statistics 64 (2): 135158.
Intergovernmental Panel on Climate Change IPCC (2007): Climate Change Impacts,
Adaptation and Vulnerability, Report of Working Group II to the Fourth Assessment Report of
the IPCC. Cambridge University Press, Cambridge: UK and New York.
IPCC. 2007. Climate Change 2007: Synthesis Report. Contribution of Working Groups I, II and
III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Ed. R.
K Pachauri and Reisinger, A. Vol. 446. November. IPCC Geneva, Switzerland.
Koop, Gary, Dale J. Poirier, and Justin L. Tobias. 2007. Bayesian Econometric Methods.
Cambridge University Press.
La Rovere, Roberto, Genti Kostandini, Abdoulaye Tahirou, John Dixon, Wilfred Mwangi, Zhe
Guo, and Marianne Bnziger (2010): Potential impact of investments in drought tolerant maize
in Africa; International Institutte of Tropical Agriculture (IITA) and International Maize and
Wheat Improvement Center (IMWIC)
Levine, R., and D. Renelt. 1992. A Sensitivity Analysis of Cross-country Growth Regressions.
The American Economic Review: 942963.
27

Lindley, D. V., and A. F. M. Smith. 1972. Bayes Estimates for the Linear Model. Journal of
the Royal Statistical Society. Series B (Methodological): 141.
Mankiw, N. G., D. Romer, and D. N. Weil. 1992. A Contribution to the Empirics of Economic
Growth. The Quarterly Journal of Economics 107 (2): 407437.
Mendelsohn, R. 2008. The Impact of Climate Change on Agriculture in Developing Countries.
Journal of Natural Resources Policy Research 1 (1): 519.
Mendelsohn, R. 2009. The Impact of Climate Change on Agriculture in Developing Countries
Journal of Natural Resources Policy Research 1: 5-19.
Mitchell, T. D., and P. D. Jones. 2005. An Improved Method of Constructing a Database of
Monthly Climate Observations and Associated High-resolution Grids. International Journal of
Climatology 25 (6): 693712.
Odusola, A.F. and A.E. Akinlo, Output, Inflation and Exchange Rate in Developing Countries:
An Application to Nigeria, The Developing Economies, XXXIX, 2001.
Odusola, A.F. and Akinlo E.A. "Food Supply and Inflation in Nigeria" International Review of
Economics and Business, Volume XLI No.8 August, 1994
Parry, M. L., O. F. Canziani, J. P. Palutikof, P. J. van der Linden, and C. E. Hanson. 2007. IPCC,
2007: Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working
Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change.
Cambridge University Press, Cambridge.
Pindyck, R. S. 2011. Fat Tails, Thin Tails, and Climate Change Policy. Review of
Environmental Economics and Policy 5 (2): 258274.
Rabassa, M., E. Skoufias, and H. G. Jacoby. 2012. Weather and Child Health in Rural Nigeria.
World Bank Policy Research Working Paper (6214). http://www-
wds.worldbank.org/servlet/WDSContentServer/WDSP/IB/2012/10/02/000158349_20121002133
547/Rendered/PDF/wps6214.pdf.
Sala-i-Martin, X. X. 1997a. I Just Ran Two Million Regressions. The American Economic
Review: 178183.
Sala-i-Martin, X. X.. 1997b. I Just Ran Four Million Regressions. National Bureau of Economic
Research. http://www.nber.org/papers/w6252.
Salvador, Barrios, Bazoumanna Ouattara and Eric Strobl (2004): The Impact of Climatic
Change on Agricultural Production: Is it different for Africa? MPRA Paper No. 6240,
http://mpra.ub.uni-muenchen.de/6240/
Savvides, A. 1995. Economic Growth in Africa. World Development 23 (3): 449458.
)#

Science Daily (2010): Climate Change One Factor in Malaria Spread, March
2010.http://www.sciencedaily.com/releases/2010/03/100303162906.htm
Tanser, F. C., B. Sharp, and D. Le Sueur. 2003. Potential Effect of Climate Change on Malaria
Transmission in Africa. The Lancet 362 (9398): 17921798.
World Bank Group. 2011. Africa Development Indicators 2011. World Bank Publications.